| Color error diffusion with error signal offsets -> Monitor Keywords |
|
|
  Color error diffusion with error signal offsetsThe Patent Description & Claims data below is from USPTO Patent Application 20080055649. Brief Patent Description - Full Patent Description - Patent Application Claims
CROSS REFERENCE TO RELATED APPLICATIONS [0001]Reference is made to commonly assigned U.S. patent application Ser. No. 10/795,011 filed Mar. 5, 2004 by Couwenhoven, et al., entitled "Multi-level Halftoning Providing Improved Texture Uniformity", and U.S. patent application Ser. No. 10/794,133 filed Mar. 5, 2004 by Couwenhoven, et al., entitled "Color Error Diffusion", the disclosures of which are herein incorporated by reference. FIELD OF THE INVENTION [0002]The present invention relates to the field of digital image processing, and more particularly to a method for digitally halftoning a continuous-tone image using error diffusion. BACKGROUND OF THE INVENTION [0003]Digital halftoning is a technique employing digital image processing to produce a halftone output image from a continuous-tone input image. In the digital halftoning technique, a continuous-tone image is sampled, for example, with a scanner or digital camera, and the samples are digitized and stored in a computer. The digitized samples (or "pixels") include discrete code values having N.sub.i possible input levels (typically N.sub.i=256, with code values ranging from 0 to 255). To reproduce this image with high quality on an output device which can print N.sub.o output levels, where N.sub.o<N.sub.i, it is necessary to produce the sensation of intermediate gray levels by suitably distributing the printed dots in the output image using some form of halftoning technique. One example of an output device requiring halftoning is an inkjet printer, which ejects a volume of ink at each output pixel to reproduce the image. The number of output levels (N.sub.o) of the printer is the number of possible volumes deposited at each pixel (including 0), and is typically in the range of 2-32. [0004]One prior art method of digital halftoning is known as error diffusion. FIG. 1 shows a block diagram describing a basic error diffusion technique. The continuous-tone input level for one pixel of the input image is shown as X. For purposes of illustration it will be assumed that the continuous-tone input level spans the range from 0 to 255. The input level X is added to a weighted error signal X.sub.fe using a summer 10 to produce a modified input level X.sub.in. (The generation of the weighted error signal X.sub.fe will be discussed shortly.) The modified input level X.sub.in for the current pixel is quantized using a quantizer 20 to form the quantized output level X.sub.o. For a binary error diffusion algorithm, the quantizer 20 will output a 0 for any input level below a threshold (typically code value 128), and a 255 for any input level above the threshold. A summer 30 receives the modified input level X.sub.in and the quantized output level X.sub.o and produces an error signal X.sub.err representing the error introduced by the quantization process. The error signal X.sub.err is multiplied by a series of error feedback weights using a weighted error generator 40 to produce the weighted error signal X.sub.fe, which is added to the continuous-tone input levels of nearby pixels which have yet to be processed using summer 10. The propagation of the errors made during the quantization process to the nearby pixels insures that the arithmetic mean of the pixel values is preserved over a local image region. [0005]FIG. 2 shows a typical set of error feedback weights used in the prior art. The error for the current pixel, located at (column,row)=(i,j) of the image, is weighted by the error feedback weights, and added to nearby pixels yet to be processed. In this example, there are four error feedback weights having the same value of 1/4, with the sum of the error feedback weights being 1. This ensures that the arithmetic mean of the image is preserved. [0006]When using the standard error diffusion algorithm to process a color image, a technique common in the prior art is to apply the error diffusion algorithm independently to each color channel. See, for example, U.S. Pat. No. 5,757,517 to Couwenhoven, et al. This arrangement is shown for processing an image with cyan (C), magenta (M), and yellow (Y) color channels in FIG. 3. This arrangement is successful at preserving the arithmetic mean of the image data in each color channel, resulting in the correct output color, but suffers from the fact that the placement of the output dots in each color channel is decorrelated, and overprints can unnecessarily occur, resulting in an undesirable noisy, grainy appearance to the printed image. To illustrate this, consider that a 16.times.16 pixel image containing a 30% tint of uniform CMY code values is to be halftoned to 2 levels (on, off) and printed on an inkjet printer. As shown in FIG. 4, the CMY color channels, when processed independently with the error diffusion algorithm, produce dot patterns that typically resemble the cyan, magenta, and yellow patterns shown as patterns 50, 60, and 70, respectively. In these patterns, the black pixels indicate locations where an ink dot will be printed, and white locations indicate white paper. When these three patterns are printed on top of each other on the page, an output pattern 80 will result. (Due to the black and white reproduction of output pattern 80 in this document, only the luminance of the pattern is shown). Each pixel in the output pattern 80 can be one of eight colors: white; C; M; Y; red (C+Y); green (C+Y); blue (C+M); or black (C+M+Y). The luminance modulation between the darker pixels (R, G, B, or black) and the lighter pixels (white, C, M, or Y) produces an undesirable grainy appearance to the human eye. Interestingly, this is not necessary, since the original input image was a 30% tint of CMY, therefore it should have been possible to produce the output image using non-overlapping patterns such that only white, C, M, or Y are produced at each pixel. The luminance modulation between these colors would be much less, providing a more pleasing result. [0007]Error diffusion methods that attempt to provide correlation between the color channels are known in the prior art. These algorithms are sometimes called "vector error diffusion" algorithms. Chapter 16 of "Digital Color Halftoning", by H. Kang (SPIE Optical Engineering Press, 1999) describes several vector error diffusion techniques. U.S. Pat. No. 5,375,002 to Kim, et al., discloses a color error diffusion method in which an error look-up table is used to help determine the color of an output pixel. U.S. Pat. No. 5,565,994 to Esbach discloses an error diffusion method in which output signals from one color channel are used to determine threshold values for other color channels to provide some decorrelation between the color channels. U.S. Pat. No. 6,637,851 to Van de Velde, et al., discloses an error diffusion method in which a luminance signal is computed and separately error diffused as an additional channel. The luminance information is then used to guide the selection of the output pixel colors. Typically, the prior art methods are computationally complex, requiring substantial additional processing power or memory. [0008]Therefore, there is a need for a color error diffusion method which provides for high quality printed images with reduced graininess, and can be implemented efficiently without requiring substantial additional computing power or memory. SUMMARY OF THE INVENTION [0009]It is an object of the present invention to provide printed color images with reduced graininess. [0010]It is a further object of the present invention to provide for improved quality of printed color images while reducing the amount of computer memory and computer processing power required to process the images. [0011]Still another object of the present invention is to provide for high quality color images when printed on an inkjet printer. [0012]Still another object of the present invention is to reduce the start-up artifacts generally associated with error diffusion algorithms. [0013]These objects are achieved by a color error diffusion method for multi-toning an input digital image having input pixels with two or more color channels, each color channel, C, having a specified number of input levels, N.sub.i, to form an output digital image having modified output levels, each color channel of the output digital image having a specified number of output levels, N.sub.o, where 2.ltoreq.N.sub.o<N.sub.i, comprising: [0014]a) determining output levels for each color channel using a process that includes a quantization step that produces intermediate error values and an error sorting step that is applied to the intermediate error values; [0015]b) determining error signals for each color channel responsive to the input levels and output levels; [0016]c) producing first error signal offset values that are a function of the input levels of the input pixel and producing shifted error signals by adding the first error signal offset values to the error signals for each color channel; [0017]d) weighting the shifted error signals for each color channel by a set of error feedback weights to determine weighted error signals for nearby pixels that have yet to be processed; [0018]e) producing additional error signal offset values that are a function of the input levels of the corresponding nearby input pixels and producing shifted weighted error signals by adding the additional error signal offset values to the weighted error signals for each color channel; [0019]f) adjusting the input levels for the nearby pixels responsive to the shifted weighted error signals; and [0020]g) repeating steps a)-f) for multiple input pixels of the input digital image to thereby provide the output digital image. Continue reading... Full patent description for Color error diffusion with error signal offsets Brief Patent Description - Full Patent Description - Patent Application Claims Click on the above for other options relating to this Color error diffusion with error signal offsets patent application. Patent Applications in related categories: 20080291499 - Multi-bit-depth error diffusion for the reproduction of color or monochrome images - An image processing system including a lookup table having multi-bit printer output levels and an imaging input level. The imaging input level may be associated with a range of input pixel densities. The system further includes a processor configured to compare the imaging input level with one or more preconfigured ... ###  How KEYWORD MONITOR works... a FREE service from FreshPatents1. Sign up (takes 30 seconds). 2. Fill in the keywords to be monitored. 3. Each week you receive an email with patent applications related to your keywords. Start now! - Receive info on patent apps like Color error diffusion with error signal offsets or other areas of interest. ### Previous Patent Application: Page coloring with color inheritance for memory pages Next Patent Application: Electrophotographic image forming apparatus and halftoning correction method thereof Industry Class: Facsimile and static presentation processing ### FreshPatents.com Support Thank you for viewing the Color error diffusion with error signal offsets patent info. IP-related news and info Results in 0.8569 seconds Other interesting Feshpatents.com categories: Tyco , Unilever , Warner-lambert , 3m |
||
